动态时间归整算法抑制前向散射直达波干扰

Suppression of Interference by Direct Arrival Wave Using Dynamic Time Warping Method

基于衍射CT成像技术,在弱散射条件下根据收发分置目标散射强度的指向分布特性,可以重建目标形状、以及周围介质的声学参数。在目标的正前方,前向散射波和直达波同时到达接收阵,特别是当目标为密度和声速接近于水的弱散射目标时,散射回波完全淹没在直达波中,抑制直达波干扰才能实现目标的识别、方位估计以及声学特性解释。基于语音识别的动态时间归整(Dynamic Time Warping,DTW)算法,是利用归整路径距离的动态规划来求解发音长短不一的孤立词模板匹配的问题。固定入射波方向,以浸没水中的无限长圆柱体目标为例获取实验数据,将DTW算法用于直达波和散射波的分离,并将消除直达波影响后的目标散射远场的强度与严格的解析结果进行了对比,结果表明DTW算法可以有效地在强直达波干扰背景中有效提取目标散射信息。

Based on diffraction CT imaging technique, shape of targets and acoustic parameters of the surrounding medium can be reconstructed by directional distribution of scattered wave in bistatic system. While in front of the target, forward scattered wave and direct wave reach the receiving array at the same time. Especially under weakly scattering condition, namely, density and sound velocity of the target are comparable to that of water, scattered echo is completely submerged in the direct wave. Suppressing interference of direct wave is important for targets identification, position estimation and interpretation of the acoustic properties. Dynamic Time Warping（DTW） is a speech recognition method, which can be used to solve template matching problems of single words with different pronunciation length. Specified incident wave direction, immersed an infinite long cylinder in water to obtain data by experiment in tank. DTW algorithm is used to separate direct and scattered waves. Scattered far-field is calculated and presents a good agreement with analytic solution. The results show that the DTW algorithm can effectively extract scattering information with strong interference of direct wave.